As illustrated in FIG. 1, systems and methods exist that allow retail outlets to use a reference speaker to demonstrate the performance of multiple demonstration speakers. For example, the Virtual Speaker Demonstration System And Virtual Noise Simulation disclosed in U.S. Pat. Nos. 7,096,169 and 7,899,656, incorporated herein by reference, enables a customer to simulate the performance of a particular demonstration speaker using a reference speaker. These virtual speaker demonstration systems simulate the output of the selected demonstration speaker by determining characteristics of the demonstration speaker and applying these characteristics to a sample acoustic input.
Significantly, however, the characteristics of a demonstration speaker are typically derived (empirically or analytically) without regard to its spatial characteristics. That is, retailers often empirically determine the characteristics of a demonstration speaker by measuring its transfer function at a single point in space. FIG. 1 illustrates an exemplary prior art embodiment, where a single microphone 102 is used to measure a transfer function of the demonstration speaker at a single point in space. Although such a transfer function may be used to simulate a demonstration speaker, the transfer function will not be an accurate representation of how the demonstration speaker would sound at any other point in space.
Audio enthusiasts understand that a key driver of sound quality in audio equipment is how sound waves disperse from a sound source. For example, acoustic phenomena such as speaker directivity, interaction between multiple drivers, and diffraction affect which frequencies of sound can be heard at different distances or angles from a speaker. Directivity, for instance, characterizes a speaker's ability to emit different spectral frequencies of an audio signal in one particular direction. In all wave-producing sources, the directivity of any source generally corresponds to the size of the source compared to the wavelengths it is generating. Thus, loudspeakers tend to radiate sound omnidirectionally (i.e., uniformly in all directions) at low frequencies, because the physical components of the speaker, such as surface dimensions and cabinetry, are generally small compared to the wavelength of the sound. However, at high frequencies, speakers tend to beam the sound, because the physical components of the speaker are no longer negligible as compared to the sound's wavelength. Thus, speakers typically generate a “beam” of high spectral frequencies directly in front of a speaker, while lower bass-like frequencies may be perceived both in front, and behind a speaker. Further complexities may be added to the speaker's spatial characteristics when multiple drivers are radiating in the same frequency range, such as the crossover region, and when acoustic diffraction occurs due to physical discontinuities, such as at the edges of the speaker cabinet. More generally, each loudspeaker radiates a different spectrum of frequencies at different angles off of its central axis (off-axis). Accordingly, each loudspeaker has a distinct transfer function that depends on the listener's location in space. Consequently, listeners located at different positions around a speaker will each hear a different spectrum of sound, even though they are each listening to the same speaker.
Directivity, and other similar spatial characteristics, may be represented as directivity patterns, illustrated in FIGS. 2A and 2B. FIG. 2A is an exemplary polar plot that illustrates the frequency gain of a demonstration speaker for a particular frequency (e.g., 10 kHz) across a 360° rotation around the demonstration speaker. As FIG. 2A illustrates, the frequency response has higher gain due to beam-forming in front of the speaker, while less energy radiates to the rear. FIG. 2B is an exemplary directivity plot illustrating how a speaker may have different frequency response gains at different frequencies, and at different angles off of the speaker's center axis. For example, a frequency of 125 Hz may have a gain of −3 dB at 90°, whereas a frequency of 1.6 kHz may have a gain of −9 dB at 90°. As these figures illustrate, the transfer function of a speaker may vary drastically as it is measured throughout different points in space. As stated above, listeners consequentially hear different spectrums of frequencies at different points in space.
Differences in the build of the speakers, such as shape, material, dimensions, and placement of components (e.g., the transducer) in the casing, may also affect the ability of the speaker to project, i.e., “throw”, a sound wave at varying distances from the speaker in a coherent fashion. As with directivity, the dispersion of the sound wave from the speaker depends on the frequency of the sound wave, the technical specifications of the speaker driver, and the dimensions (e.g., shape size, positioning, etc.) of the speaker horn. As a result, audio enthusiasts generally understand that two different speakers may exhibit different sound quality as a result of how sound waves disperse from the speakers, i.e., the spatial characteristics of the speakers.
Although audio retailers are generally aware that different speakers disperse sound differently (reflecting a difference in sound quality), audio retailers typically do not provide an effective way to demonstrate this difference to audio consumers. As noted above, prior demonstration systems such as those disclosed in U.S. Pat. Nos. 7,096,169 and 7,899,656 and implemented in Crutchfield's Virtual Speaker System, are generally limited to the acoustic performance of speakers as measured from a single point in space. As FIG. 1 illustrates, acoustic characteristics (e.g., frequency responses and transfer functions) of a speaker are typically measured from a single point in space, 102, in relation to the speaker. This single data point is then applied to a reference speaker 106 to simulate the performance of the demonstration speaker 101. Accordingly, demonstration systems currently provided by audio retailers typically do not demonstrate the spatial performance of a speaker, i.e., the impact of sound quality due to the dispersion of the sound waves from the speaker, because the acoustic characteristics are only measured from a single point in space.
For one, audio retailers do not have an efficient means for empirically measuring spatial characteristics at multiple points in space. Typically, determining directivity pattern is time and resource intensive, usually requiring a retailer to make discrete sequential measurements of a speaker's transfer function at different angles around speaker. While systems such as microphone arrays exist, which enable retailers to make several simultaneous measurements of a speaker's transfer function in space, several drawbacks exist. Microphone arrays are generally applied to problems in two categories of acoustics: 1) beam-forming; and 2) near-field acoustic holography (“NAH”). Beam-forming microphone arrays process the microphone signals in a way that causes the array system to be more sensitive to sound coming from one particular direction. NAH is concerned with using acoustic measurements to determine the vibration of an acoustically radiating surface. Significantly, these applications of microphone arrays are not typically designed to capture the spatial performance characteristics of a loudspeaker.
Even where audio retailers have spatial characteristics measured from multiple points in space, problems exist in combining speaker characteristics to effectively simulate these spatial characteristics to users. Electro-acoustic modeling software packages (such as EASE®, MEYER®, or ANSYS®) provide analytic acoustic information (e.g., directivity and dispersion). However, these modeling software packages are generally time and resource-intensive to use. Accordingly, an efficient means for measuring the acoustic characteristics of any sound source, and simulating the sound source with empirical data is needed.
Further problems exist with combining different empirically derived spatial characteristics. When combining and processing characteristics from multiple sources at different points in space, the oscillatory behavior, thus the complex nature of the data, must be carefully considered to avoid introducing unrealistic distortions. Accordingly, an efficient means for combining the acoustic characteristics measurements of sound sources while minimizing the introduction of distortions is needed. Other problems and drawbacks also exist.
Even when the empirically derived spatial characteristics are suitably combined, simulation errors remain due to the single degree of freedom (SDOF) afforded by the traditional reference speaker. The remaining error is due to the difference in acoustic spatial radiation patterns between the target simulation speaker and the reference speaker used to perform the simulation. Thus, there exists a need for a multi-degree of freedom (MDOF) reference speaker, such as a speaker array consisting of several acoustic drivers that may allow the control over the reference speaker's spatial radiation pattern, and in turn, allow accurate simulation of the target simulation speaker's temporal and spatial characteristics simultaneously.